A runtime resolution scheme for priority boost conflict in implicit coscheduling

Abstract

High-performance parallel and scientific applications are composed of multiple processes running on distinct CPUs that communicate frequently. Due to the synchronization needs of such applications, performance is greatly hampered if their processes are not scheduled simultaneously on the CPUs. Implicit coscheduling (ICS) is a well-known technique to address this problem in multi-programmed clusters, however, traditional ICS schemes do not incorporate steps to adequately deal with priority boost conflicts, leading to significantly degraded performance. In this paper, we propose the use of runtime difference in contention across nodes to provide more sophisticated coscheduling decisions in response to the conflicts. We also present a novel coscheduling scheme termed PROC (Process ReOrderingbased Coscheduling) that adaptively regulates the scheduling sequence of conflicting processes based on the rescheduling latency of their correspondents in remote nodes. We perform extensive simulation-based experiments using both synthetic and realistic workloads to analyze the performance of PROC compared to alternatives such as local scheduling, a widely used batch scheduling, gang scheduling, and existing ICS schemes. The results show that all ICS schemes commonly experience priority boost conflicts, and that the proposed PROC significantly outperforms other ICS alternatives (or batch scheduling) by up to 50.4% (or 72.5%) in the average job response 2 J.-L. Yu et al. time. This improvement is achieved by reducing wasted idle time and spinning time without sacrificing fairness.

Design and Implementation of a Contention-Aware Coscheduling Strategy on Multi-Programmed Heterogeneous Clusters

“…According to these results, we can conclude that the proposed CC approach is very promising, especially for the hunting for wasted idle cycles in clusters, improving the overall system utilization, and this is in good agreement with the simulation results reported in [6].…”

“…To address this problem, in our previous study [6], we proposed a flexible CDC scheme that aims to efficiently resolve priority boost conflicts, and showed its feasibility on multi-programmed clusters. In this paper, we exhaustively explore the design issues and implementation details of our Contention-aware Coscheduling (CC) scheme over Myrinet [18]-based high performance Linux clusters.…”

Empirical performance evaluation of schedulers for cluster of workstations

“…It has been ensured that execution time of each job must be nearly equal. During the workload generation arrival time, execution times are characterized to fit accurate workload model [23,24]. Our workload is then completed by the eight scientific computing i.e.…”